Perplexity Comet Voice Mode 2 — powered by GPT Realtime 1.5
Comet upgraded Voice Mode to GPT Realtime 1.5 with 25%+ better reliability, screen awareness and voice browser navigation. Rolling out on desktop and Android.
Categories: OpenAI, Google, Perplexity, Vercel, Notion (AI features)
10 noviniek
Results
Pondelok 22. júna 2026
Nedeľa 21. júna 2026
Vercel changelog: Sandbox up to 24h, Passport public beta, Connect for agents, GLM 5.2 on AI Gateway
Alongside its eve framework Vercel shipped infra for long-running agents. Sandbox now runs up to 24 hours, Passport is in public beta, Connect gives agents secure access to external services, and AI Gateway added GLM 5.2.
Gemini API changelog: TTS streaming for flash-tts-preview, Veo and Imagen models scheduled for shutdown
Google added speech streaming for `gemini-3.1-flash-tts-preview` and announced shutdown dates for legacy media models. Veo retires June 30, Imagen 4 and earlier Gemini image models on August 17, 2026.
Introducing Record & Replay for Codex
Codex on macOS records a workflow you demonstrate once and converts it into a reusable skill it can execute autonomously on demand.
Notion 7.22: simple table cell merging, workspace API rate limits and Workers runtime
Notion 7.22.0 ships simple-table cell merging, workspace-level API rate limits scaled to plan, and Workers — a hosted sandbox runtime for custom code that extends Notion without managing your own servers.
Sobota 20. júna 2026
Introducing OpenAI for Healthcare and improved health intelligence in ChatGPT
OpenAI launched OpenAI for Healthcare — a product suite for healthcare organizations with HIPAA support — and improved health intelligence in ChatGPT, which now sees 230M+ weekly health-related queries.
Piatok 19. júna 2026
OpenAI frontier models and Codex are now available on AWS
OpenAI made GPT-5.5, GPT-5.4 and Codex available via Amazon Bedrock — ending Azure's hyperscaler exclusivity for OpenAI APIs.
Introducing eve — open-source agent framework
Eve is a TypeScript-native open-source framework where every agent = a directory of files (instructions.md + tools/ + skills/). Built in: durable execution, per-agent sandbox, human approvals, evals, OpenTelemetry. Plus the Sandbox runtime now goes up to 24h.
OpenAI Codex Record & Replay — show-don't-tell workflow capture
Codex app v26.616 captures a workflow from a single demonstration and generates a natural-language skill description that generalizes to new inputs. Requires Computer Use.
OpenAI frontier models and Codex are now available on AWS
OpenAI made its frontier models and Codex available on AWS — ending Azure's de facto exclusivity for OpenAI API workloads.